In our search for the possible causes of obesity in the Pima Indians, we have previously shown using indirect calorimetry (respiratory chamber and venti-lated hoods), that metabolic rate varies between people more than can be explained by individual differences in body size and composition, age and sex. This variability also seems partially genetically determined. More importantly, we have shown that subjects with a low metabolic rate are at greater risk of gaining body weight. We have therefore, investigated the possible causes of the variability of metabolic rate. Our studies have shown that: 1) females have lower metabolic rates than males even after adjusting for differences in body size and composition; 2) age is not a major deter- minant of sedentary energy expenditure in sedentary conditions; 3) physical fitness does not change energy expenditure beyond the difference in body weight and composition; 4) the sympathetic nervous system (SNS) activity is a determinant of metabolic rate in Caucasians, but not in the Pima Indians, suggesting that the decrease in SNS-mediated energy expenditure in Pima Indians may predispose them to obesity; 5) metabolic rate seems to be regu- lated to maintain a genetically determined body temperature. Studies in 10-year old children have shown that Pima Indians are much heavier and fatter than Caucasians, but their metabolic rate is normal for their body size and composition. Studies using the doubly-labeled water method to measure energy expenditure in free-living conditions have shown; 1) the variability in energy expenditure is much larger in free-living conditions than in a respira- tory chamber suggesting large differences in physical activity; 2) females have lower levels of physical activity; 3) overweight subjects have lower physical activity than lean subjects; 4) Intensive training does not result in metabolic adaptations lowering energy requirements in female athletes.